raindancewx

Globally, the Summer heat waves this year have been quite consistent with the warming of Nino 3.4. Western Canada, Mexico/Caribbean states, Western Europe, the Maghreb/North Africa/Middle East, India/Pakistan, Western China, with the US relatively unaffected except immediately downwind of the Canadian hot spots. Southwestern Canada had temperatures near 100F in May, with Pakistan seeing near 120F recently. Mexico got to the mid-80s as high up as 8,000 feet on the Central Plateau against average highs in the mid-70s. 2003 and 2015 are some recent years when El Nino conditions were present by Summer, with major heat waves across Europe and the Middle East. Here is 2015 in Europe. I believe August saw temps over 120F for a time in Iraq. The C to F conversion is (C x 1.8) + 32. So the 35C-40C range is 95F to 104F. From this account: extreme temps on twitter. 110s in Western China - (46.6 x 1.8)+32 = 116F (+15F v. average high for June)

I entered the long Tip response into an AI bot with the following prompt - "please explain this so a five year old would understand". Here is the translation for everyone who hates jargon and poor writing: Okay, so imagine there's a big puzzle in front of us that we are trying to put together. We have different pieces of information that fit together like the pieces of a puzzle. But right now, some of those pieces aren't fitting together very well because something is interfering with them. This interference is caused by something called GLAAM, which makes the air move in a certain way in the atmosphere. Right now, the GLAAM is not doing a very good job and it's making it harder for us to figure out how things are working together. There are other things going on too, like ocean patterns and wind patterns that can also affect how everything fits together. And all of these things can make it hard for us to predict what's going to happen in the future. It's like trying to solve a really hard puzzle, but some of the pieces don't seem to match up like they should.

Remember how DT was freaking out about the PDO reversing very quickly? Complete bullshit for May - Keep in mind, these values are now "normalized" from the "un-normalized" values Mantua produced before. It's kind of dumb to have a normalized index from something that's produced with such tortured methodology in the first place. But it is what it is. Behaviorally, I still like 1972, it's been near identical on both PDO values and Nino 3.4 values all year. It followed a multi-year La Nina, and the 1971-72 to 1972-73 transition is one of the only La Nina to El Nino transitions in the last century where the La Nina was cold in the West. 2022-11-01T00:00:00Z -2.12 2022-12-01T00:00:00Z -1.98 2023-01-01T00:00:00Z -0.92 2023-02-01T00:00:00Z -1.1 2023-03-01T00:00:00Z -1.63 2023-04-01T00:00:00Z -2.18 2023-05-01T00:00:00Z -1.65 https://oceanview.pfeg.noaa.gov/erddap/tabledap/cciea_OC_PDO.htmlTable?time,PDO Previous developing El Nino May PDO (normalized) values: 2018 - 0.00 2014 - +1.41 2009 - -0.75 2006 - +0.35 2002 - -0.82 1997 - +1.84 1994 - +0.83 1991 - -1.04 1986 - +1.06 1982 - -0.60 1976 - -0.73 1972 - -1.81 1968 - 0.70 1965 - -0.48 1963 - -1.07 1957 - +0.14 1951 - -0.91

This is how I see the strength of this event playing out. The Euro has some skill at this range for winter, but the result is almost always at the edge of what it forecasts.

Solar above 55 sunspots annualized from July-June in Albuquerque and a lot of the high terrain in the West has statistically significantly differences in snowfall tendencies in March. Here, we're 1/36 for March snow over 3 inches in low solar years since March 1932. We're 16/56 I believe in high solar March for over 3 inches of snow. It's literally 10:1 frequency. I always test solar stuff with hypothesis testing, assuming no difference. But 10:1 passes at highly statistically significant levels using a difference in proportions test - but it only works in March. One of the things about ENSO that is lost is why it does what it does. The deserts of the world are models of strong highs and consistent weather. But dry places that are high up are unbalanced. You can see locally we're far more likely to see extreme temperatures on the cold side of average than the hot side of average at all times of year. El Nino is much more likely to do that for highs via clouds, rain and snow. But enough snow in the winter does the job here too. It gets very cold here when snow sticks around for a few days when the dew points crater back down. These are literally counts of days that are 5 to 25 below average, and 5 to 25 above average in 1991-2020 in Albuquerque, against 1991-2020 averages by month. You can see we are almost never 15F above average for highs or lows.

The 90 degree timing doesn't matter for anything locally in El Nino. In La Nina it's the opposite of what you think. Late heat onset is tied to warmer winters in the Southwest. Early heat onset is tied to colder winters. We hit 90F May 7th in both 2020-21 and 2022-23. Both years saw pretty intense and severe cold patterns at times in the West, with similar behavior in other early onset years for the heat. Intense early heat for the West almost always ties into the MJO being strong in the right phases for the West in winter. Right now we have: +El Nino following La Nina +Volcanic effects that are opposite normal (extra water vapor/heating) +High solar That's it. High solar with El Nino is a very reliable signal for heavy snow in the Southwest. El Nino after La Nina tends to be cold/wet or both int he Southwest. Volcanic years tend to be very cold with unusually intense storms nationally in the Fall/Spring. I will say this - Canada and Russia have been very hot at times in recent weeks. I doubt the Summer heat will be interrupted by nice early cold shots from the North if that continues. 38C (100F+) in Siberia at 52N when we can't buy a 90F day in the high desert at 35N is pretty impressive. There is a part of me that thinks of a lot of the high-terrain of Northern Mexico and then the SW US will eventually get expanded into "subtropical highland" climate ala Mexico City as things continue to warm - but we'll see.

CPC has 28.33C for May 2023 now. (I did adjust the early 2023 numbers to match CPC's edits - they constantly tinker with the data). https://origin.cpc.ncep.noaa.gov/products/analysis_monitoring/ensostuff/detrend.nino34.ascii.txt I have no idea if the winter will look like 1972-73. But I'm increasingly on-board with that year as a good estimate for how warm Nino 3.4 gets. I can see a month or two around 28.5C in the Fall, but only 28.0C or so in the actual winter. Nino 3.4 on the weeklies already looks like it may have peaked at 28.4C. Following the foot steps of June 1972 (28.18C) is doable. I've been pleased with the Summer so far locally. No 90s yet. None coming either in the near term. Outside chance we reach 6/10 or later without reaching 90 here - hasn't been done in decades (1999). Follows a cold season with 111 lows that were freezing or colder. I'm very curious to see how the transitional seasons go this coming cold season. My current guess is another wild Fall.

1994 is actually the top matching May for developing El Nino in the past 100 years or so.

The Canadian has a cold winter for the Southern US. Still no +PDO depicted. Warm tongue east of Japan, some cold off the Western part of North America. I actually think the look below is pretty defensible as a realistic outcome. I'd expect a stormy, but not real cold winter for the NE/NW US in the look below.

Also, subsurface heat content is not really rapidly increasing. Nino 3.4 is going to finish May in the low 28C range. Average years see a sea surface temperature drop from May to Dec-Feb. But El Ninos (usually) still see actual SSTs fall from May to the following DJF. No year has even seen more than a 0.4C rise from May to DJF. Mexican precipitation and temperature patterns are more strongly correlated to ENSO this time of year than the US. May had canonical El Nino temperatures in place there spatially. The map has negative correlations for Northern Mexico. So +Nino 3.4 = cold. Precipitation is still kind of doing its own thing - not really a dry south / wet north look.

CFS has an unusual look for June for the US temperature profile. Not common in El Nino from some quick looks. I was able to get a decent match by warming up a blend of 1963, 1992, with -2018 added in. Ignore the color template scale - just pretend its -7 to +7 - I made the blend much warmer. The actual pattern is essentially a -3 to +3 look. I do intend to list the analog model I use for global conditions for purchase or at least download somewhere. Since I have annualized solar conditions there for July-June, I'm going to wait until June is over before I release it.

I've been thinking about listing my analog model online for purchase online. Would any of you be interested? You basically type in your expected winter outcomes on seven global variables. Then all years back to 1931 are scored based on a weighted statistical model I developed off of a multi-factor polynomial regression. The scoring is then sortable, and I have a listing of what each variable corresponds to over the long-term (rain, snow, temps, etc in winter). Since ENSO is important but not dominant in the model, you can get strong scoring matches for ENSO years that are not in the same third of the scale (i.e. a rogue La Nina or Neutral might match this year) in special circumstances. Usually there are one or two odd ducklings in the seven variable set. So I pick a fixer year among the medium-level match years to bring the pack to the point of mimicking actual global conditions. The main purpose of the tool is to systematically score years as similar or dissimilar without needing to run through hundreds of maps manually. Knowing the most dissimilar years can actually be useful too, to see if reversing the exact opposite years produces similar matches to the most similar years. The objective scoring last year had both 1984 and 2016 as strong matches as an example. Pretty cold and severely wet winters in the West, just like we actually saw. 1995 was in there as among the least similar La Ninas, which is part of why I was pretty bullish on a shitty season for snow for most of you in the Northeast.

My current working blend for Summer, at least for precipitation is May 1972/1997. I think maybe 1972/2012 is better for temperatures, but haven't really looked at that yet. Here is that blend in May v. May 2023 to date. You can see the wet spots are similar running up the front range, NE, and in the SW, with the dry spots similarly placed too. Keep in mind, I'm incorporating this verifying in some form by month end -

I don't think anyone is actually dumb enough to forecast seasonally just based on ENSO. I have a seven factor method for seasonal stuff nationally. Then once the general idea is determined, I try to narrow it on the monthly patterns. But that's much harder. The atmospheric river thing in La Nina is not really new though either. The Great Flood of 1862 in the West was not an El Nino as an example. These are not new patterns. https://en.wikipedia.org/wiki/Great_Flood_of_1862 Any of this sound familiar? The Great Flood of 1862 was the largest flood in the recorded history of California, Oregon, and Nevada, inundating the western United States and portions of British Columbia and Mexico. It was preceded by weeks of continuous rains and snows that began in Oregon in November 1861 and continued into January 1862. This was followed by a record amount of rain from January 9–12, and contributed to a flood that extended from the Columbia River southward in western Oregon, and through California to San Diego, and extended as far inland as Idaho in the Washington Territory, Nevada and Utah in the Utah Territory, and Arizona in the western New Mexico Territory. The event dumped an equivalent of 10 feet (3.0 m) of water in California, in the form of rain and snow, over a period of 43 days.[3][4] Immense snowfalls in the mountains of far western North America caused more flooding in Idaho, Arizona, New Mexico, as well as in Baja California and Sonora, Mexico the following spring and summer, as the snow melted. ...The weather pattern that caused this flood was not from an El Niño-type event, and from the existing Army and private weather records, it has been determined that the polar jet stream was to the north, as the Pacific Northwest experienced a mild rainy pattern for the first half of December 1861. In 2012, hydrologists and meteorologists concluded that the precipitation was likely caused by a series of atmospheric rivers that hit the Western United States along the entire West Coast, from Oregon to Southern California.[6] An atmospheric river is a wind-borne, deep layer of water vapor with origins in the tropics, extending from the surface to high altitudes, often above 10,000 feet, and concentrated into a relatively narrow band, typically about 400 to 600 kilometres (250 to 370 mi) wide, usually running ahead of a frontal boundary, or merging into it.[7][8] With the right dynamics in place to provide lift, an atmospheric river can produce astonishing amounts of precipitation, especially if it stalls over an area for any length of time. ...The floods followed a 20-year-long drought.[9] During November, prior to the flooding, Oregon had steady but heavier-than-normal rainfall, with heavier snow in the mountains.[10]: 76–83  Researchers believe the jet stream had slipped south, accompanied by freezing conditions reported at Oregon stations by December 25. Heavy rainfall began falling in California as the longwave trough moved south over the state, remaining there until the end of January 1862, causing precipitation to fall everywhere in the state for nearly 40 days. Eventually, the trough moved even further south, causing snow to fall in the Central Valley and surrounding mountain ranges (15 feet of snow in the Sierra Nevada).[11][12]

The Jamstec seems to have a lower peak than a lot of the other models. Also has the transition away from an east-based event in Fall. I've mentioned that as a fairly likely outcome. For now, my assumption is an early peak in Nino 3.4. Halloween, plus or minus 30 days. Slow weakening after. The 1972-73 event weakened quickly despite a very healthy peak. Part of that is just how many La Ninas surrounded it: 1970-71, 1971-72, 1973-74, 1974-75, 1975-76. Never really had much of a chance. We've had multiple multi-year warm ENSO year events recently. I don't buy three initializing El Nino events in a row lasting two years each initialization. The image showing Nino 3.4 falling below +0.5C by winter 2024-25 matches my expectations. But I'd expect to be on the low side of the envelope starting around Jan-Feb.

We're still running neck and neck with the great El Nino of 2017. Nino1+2 Nino3 Nino34 Nino4 Week SST SSTA SST SSTA SST SSTA SST SSTA 10MAY2017 25.3 0.7 27.7 0.4 28.3 0.4 28.9 0.2 10MAY2023 27.0 2.4 28.2 0.9 28.3 0.5 29.0 0.2 In all seriousness though, we're in May now. Nino 3.4 has never warmed more than 0.4C from May to DJF in any year since 1950. Here are the similar Jan-Apr Nino 3.4 years I listed awhile ago in May. 1957: 28.55C - we're running behind 1963: 27.63C - we're warmer 1972: 28.32C - continues to be dead on 1997: 28.58C - behind 2014: 28.25C - very close I'm increasingly convinced 28.0C for winter 3.4 (+1.5C v. 1951-2010) is the right ballpark rather than the 28.5C - 29.0C the models try to show. In the modern climate, Nino 3.4 averages around 27.9C in May and falls off by over 1.3C or so from May to Winter. So you essentially need the ONI (difference to average) to grow by over 1.3C to even offset the normal trend. If you use 2014 as an example, it fell off from 28.25C in May to 27.19C in DJF, which is pretty close to the normal 1.3C fall off. The 1997 event grew to 28.87C by DJF, one of the biggest gains ever, but only 0.3C. 1963 fell off by raw temps. 1972 was the same in May and DJF. 1957-58 cooled off 0.4C or so from May. I'd expect DJF to be around 0.4C less than whatever May finishes at in Nino 3.4, give or take 0.2C or 0.3C. We'll see though.

I'm always amused by the talk of ENSO losing effectiveness over time as a driving factor. The truth is, it was never particularly important as a driver of temperature for most of you. I talk about it, because it is actually important where I am. But it's not really relevant in the Northeast really. PDO is more predictive for temperatures by a country mile in the NE US. In the deserts, you get phenomena if you really look through the records. In the high/colder desert/steppe land, the increase in cloud cover, snow and rain is correlated to lowering highs much more than lows in El Nino. In the warmer deserts at higher lattitudes, this is also true, but the effect is much more pronounced in Spring when there is hardly any rain normally. In La Ninas, you tend to have colder nights in the deserts as the dry air is even drier. Also fewer than average cloudy nights. As an example, we had well over 110 lows 32F or colder for Oct-May 2023, most since 2009-10, and well above the 30-year average of 92. It was 21F as recently as early April with dew points near 0F. When I look at only El Nino years, the biggest y/y changes in Nino 3.4 account for anywhere from 0.25 - 0.55 r-squared for changes in winter highs in the high terrain of NM and also old Mexico, parts of AZ and TX.

Nino 1.2 leads changes in the PDO. For now I still like the PDO to be negative in the winter. But of course it's going to revert toward less negative conditions. I just noted it was third lowest in almost 100 years. It's not going to stay that severe. The entire north Pacific could warm. Without the cold tongue east of Japan, that's still a neutral or negative PDO. The PDO has a tendency to behave differently in the high-sun half of the year v. the low-sun half of the year anyway. It could easily flip very positive by winter, but won't be clear until October. The interesting thing with 1972-73 is it kept flipping positive/negative within the cold season. Part of why that's such an interesting winter. I generally think of Nino 4 as the immediate connection to the PDO. So Nino 1.2 eventually becomes Nino 4 / PDO changes because it has to spread out. But right now, Nino 4 is only slightly warm, with no immediate pressure from severe warming. The 2020-21, 2021-22, and 2022-23 winters saw very cold Nino 4 readings - coldest in over a decade at times - and so yes, the PDO went severely negative. 1972-11-01T00:00:00Z 0.26 1972-12-01T00:00:00Z -0.09 1973-01-01T00:00:00Z 0.18 1973-02-01T00:00:00Z -0.07 1973-03-01T00:00:00Z -0.32 1973-04-01T00:00:00Z -0.74

Nov-Apr PDO finished at -1.65 on the JISAO/Mantua index. That's third most negative back to 1931-32 for a Nov-Apr. These are your -1 to -2 PDO Nov-Apr years. 1971-72, 1948-49, 1961-62, 1949-50, 1990-91, 1970-71, 1975-76, 2008-09, 1950-51, 2011-12, 1973-74, 1956-57. Of those, 1971, 1990, 1975, 2008, 1950, 1956 turned into El Ninos. These are your "El Nino following major -PDO" years then. 1951-52, 1957-58, 1972-73, 1976-77, 1991-92, 2009-10. Actually not a bad match for April, which is a good sign. That blend is actually severely cold in most of the US in the Fall, but then it lets up December. Would be a pretty good winter in the Southern US, but not particularly cold anywhere in winter itself. There is some tendency for severe cold in the Fall in volcanic years - this might actually be a decent blend for now.

1995-96 is interesting as a near perfect anti-log, at least conceptually: - First full recovery from Pinatubo (2023 is still seeing volcanic distortions) - East-based La Nina (east based El Nino) - Positive PDO (Negative PDO) - Followed an El Nino (followed a La Nina) - Near solar minimum (near solar maximum) - Weak ENSO (strong ENSO)

One weird thing about this sequence is that the Modoki La Nina and traditional El Ninos often both favor the West for heavy snow. I wouldn't necessarily expect massively different snowfall patterns in a traditional El Nino - I'd look for good snows for a lot of the West, with fluky heavy snows in the South. I thought the NE would do alright for snow by Northern New England last year. But this year, you'd probably just see more fluky snows in the East generally. If you look at snow in the five years I listed (1997-98, 1991-92, 1972-73) minus (1957-58, 2014-15), it's likely quite similar. If we have a strong El Nino, that is traditional (not a Modoki and east based), with a -PDO, then the obvious years to subtract out are 2014-15 and 1957-58 which are super +PDO events and pretty snowy in the Northeast. The blend for this winter is going to need three unusual/rare components: - Huge warm up year over year in DJF (+2.0C y/y is doable - and very rare) - Volcanic winter - PDO & ENSO in opposite phases These are your "cold to warm" by +2C or greater warm up years for Nino 3.4 -

Conceptually, the Canadian look for winter is close to 1972-73, 1991-92, 1997-98 minus 1957-58, 2014-15. I did warm that blend up by 0.2C since it is centered on the mid-1980s. PDO, ENSO, and IOD are in the right phases, with the Atlantic/North Pacific not super far off. Big heat east of Japan, at 60E, 0N, El Nino centered at 120W, etc. The paper on the IOD has the right idea. I'm always a bit skeptical about the lead time, especially since the Indian Ocean seems to be warming faster than the other oceans. But the 14 month lead idea is consistent with an 11/2023 peak for this event.

I didn't run the absolute value calculation each year. But I suspect these are the closest Nino 3.4 transitions since 1950 for Jan-Apr to observations in 2023. 1957 26.04 26.54 27.46 28.23 28.55 28.36 28.17 27.69 27.44 27.42 27.62 27.90 1963 25.77 26.22 27.18 27.78 27.63 27.62 27.78 27.48 27.40 27.36 27.47 27.62 1972 25.62 26.30 27.09 27.89 28.32 28.18 28.14 27.95 27.95 28.26 28.61 28.69 1997 26.01 26.38 27.04 27.98 28.58 28.82 28.86 28.75 28.85 29.08 29.12 28.89 2014 26.05 26.14 27.00 27.90 28.25 27.96 27.23 26.82 27.01 27.16 27.46 27.31 Blend 25.90 26.32 27.15 27.96 2023 25.83 26.29 27.18 27.91 -99.99 -99.99 -99.99 -99.99 -99.99 -99.99 -99.99 -99.99 Those five years would be a pretty severe cold season, outside of a major warm up in January. November and March (kind of like this past year really) would likely be quite severe. The severe fall/spring should play out again in some form with the volcanic activity contributing to that. March 1973 and March 2015 have some really cold storms in the West super far south, with snow down to the suburbs of Guadalajara in March 2015, and historic late season snows throughout New Mexico in 1973. The DJF blend of the five is: 1957-58 (28.15C, +1.65C against 60-year means in DJF in Nino 3.4) 1963-64 (27.36C, +0.86C) 1972-73 (28.33C, +1.83C) 1997-98 (28.87C, +2.37C) 2014-15 (27.18C, +0.68C) Blend: (27.98C +1.48C)

It's been tough finding good matches at this time of year with the SST configuration. But 1972/2012 as a blend has been decent for the main ENSO zones. That's basically a cooler version of what the Canadian shows too, right? -PDO, core of the El Nino centered at 120W.

The great sort-of east-based El Nino of 2023-24. Up to date subsurface readings (Feb, Mar, Apr) for 100-180W, at 0-300m below the surface: 2023: +0.09 / +0.84 / +1.19 ------------------------------- 2014: +0.39 / +1.60 / +1.41 2018: -0.11 / +0.51 / +0.80 2018: -0.11 / +0.51 / +0.80 ------------------------------- Blend: +0.06 / +0.87 / +1.00 Not a lot of similar years for this Feb-Apr for the subsurface progression. Actual weather in May is forecast to be much colder than that blend too. April 1997, 2014, 2015 are the only Aprils since 1979 with a warmer subsurface for 100-180W than April 2023. The cold waters at depth in the 100-180W zone, on the west side of the image, imply an early peak and then steady weakening after the Fall - but we'll see. I don't really have an issue with this getting pretty healthy for a while. Not sold on it lasting through winter though at anything near historical strength though.